Bio-inspired computation and its applications in image processing: An overview

Abstract

Almost all design problems in the sciences and engineering can be formulated as optimization problems, and many image processing problems can also be related to or formulated as optimization problems. These optimization problems can be solved by optimization techniques. However, these problems are often highly nonlinear and are subject to multiple nonlinear constraints, which makes them very challenging to solve. The further complication to these challenges is the stringent time requirements and high dimensionality, which means that traditional optimization techniques, such as gradient-based methods cannot deal with such kinds of problems well. Recent trends tend to use bio-inspired optimization techniques as a promising alternative, and it is usually combined with traditional methods, especially in the area of image processing. These bio-inspired computational methods are usually based on swarm intelligence and can be very effective in coping with nonlinearity in real-world problems. This chapter presents an overview of bio-inspired computation and its application in image processing, including some current trends and important issues, such as efficiency and time constraints.